22.01.2014 Views

Metatheory - University of Cambridge

Metatheory - University of Cambridge

Metatheory - University of Cambridge

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

5. Soundness 38<br />

Again, it is easy to check that line 4 is shiny. As is every line in this TFL-pro<strong>of</strong>.<br />

I want to show that this is no coincidence. That is, I want to prove:<br />

Lemma 5.2. Every line <strong>of</strong> every TFL-pro<strong>of</strong> is shiny.<br />

Then we will know that we have never gone astray, on any line <strong>of</strong> a pro<strong>of</strong>.<br />

Indeed, given Lemma 5.2, it will be easy to prove the Soundness Theorem:<br />

Pro<strong>of</strong> <strong>of</strong> Soundness Theorem, given Lemma 5.2. Suppose Γ ⊢ C . Then there<br />

is a TFL-pro<strong>of</strong>, with C appearing on its last line, whose only undischarged<br />

assumptions are among Γ. Lemma 5.2 tells us that every line on every TFLpro<strong>of</strong><br />

is shiny. So this last line is shiny, i.e. Γ ⊨ C .<br />

■<br />

It remains to prove Lemma 5.2.<br />

To do this, we observe that every line <strong>of</strong> any TFL-pro<strong>of</strong> is obtained by<br />

applying some rule. So what I want to show is that no application <strong>of</strong> a rule<br />

<strong>of</strong> TFL’s pro<strong>of</strong> system will lead us astray. More precisely, say that a rule <strong>of</strong><br />

inference is rule-sound iff for all TFL-pro<strong>of</strong>s, if we obtain a line on a TFLpro<strong>of</strong><br />

by applying that rule, and every earlier line in the TFL-pro<strong>of</strong> is shiny,<br />

then our new line is also shiny. What I need to show is that every rule in TFL’s<br />

pro<strong>of</strong> system is rule-sound.<br />

I shall do this in the next section. But having demonstrated the rulesoundness<br />

<strong>of</strong> every rule, Lemma 5.2 will follow immediately:<br />

Pro<strong>of</strong> <strong>of</strong> Lemma 5.2, given that every rule is rule-sound. Fix any line, line n,<br />

on any TFL-pro<strong>of</strong>. The sentence written on line n must be obtained using a<br />

formal inference rule which is rule-sound. This is to say that, if every earlier<br />

line is shiny, then line n itself is shiny. Hence, by strong induction on the length<br />

<strong>of</strong> TFL-pro<strong>of</strong>s, every line <strong>of</strong> every TFL-pro<strong>of</strong> is shiny.<br />

■<br />

Note that this pro<strong>of</strong> appeals to a principle <strong>of</strong> strong induction on the length <strong>of</strong><br />

TFL-pro<strong>of</strong>s. This is the first time we have seen that principle, and you should<br />

pause to confirm that it is, indeed, justified. (Hint: compare the justification<br />

for the Principle <strong>of</strong> Strong Induction on Length <strong>of</strong> sentences from §1.4.)<br />

5.2 Checking each rule<br />

It remains to show that every rule is rule-sound. This is actually much easier<br />

than many <strong>of</strong> the pro<strong>of</strong>s in earlier chapters. But it is also time-consuming, since<br />

we need to check each rule individually, and TFL’s pro<strong>of</strong> system has plenty<br />

<strong>of</strong> rules! To speed up the process marginally, I shall introduce a convenient<br />

abbreviation: ‘∆ i ’ will abbreviate the assumptions (if any) on which line i<br />

depends in our TFL-pro<strong>of</strong> (context will indicate which TFL-pro<strong>of</strong> I have in<br />

mind).<br />

Lemma 5.3. Introducing an assumption is rule-sound.<br />

Pro<strong>of</strong>. If A is introduced as an assumption on line n, then A is among ∆ n ,<br />

and so ∆ n ⊨ A.<br />

■<br />

Lemma 5.4. ∧I is rule-sound.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!